Microchip’s FPGA stack includes real-time Ethernet protocols

Microchip Technology has released PolarFire FPGA and SoC solution stacks for smart robotics and medical imaging. These new releases build upon Microchip’s smart embedded vision, industrial edge, and intelligent edge communications stacks already available.

 

The solution stacks include firmware and IP cores for AI-assisted 4K60 computer vision, a diverse set of ready-to-use sensor and camera interfaces, and integrated hardware for high-speed Ethernet protocols. Real-time ROS-2 compatible cores facilitate robotics tasks for perception and coordinate transformation. The stacks offer time-sensitive industrial networking protocols for OPC/UA, rich operating systems support,t and asymmetric processing commonly used in industrial automation. Software design kits allow for a high-level of customization and support diverse development environments centered around C/C++, RTL, and popular machine learning frameworks, including the SmartHLS IDE, VectorBlox Accelerator SDK, and the Libero SoC Design Suite which has been certified for applications needing IEC61503 SIL 3 functional safety. The solution stacks bring together the industry’s most power-efficient and secure mid-range PolarFire FPGAs and PolarFire SoC FPGAs, a rich mix of hardware and software solutions with cyber security protections that allow system designers the freedom to innovate in medical imaging and robotics applications.

 

One example of  Microchip’s innovative solution stacks can be found in its recently announced PolarFire FPGA Ethernet Sensor Bridge that works with the NVIDIA Holoscan sensor processing platform. With its ability to bridge real-time sensor data to NVIDIA Holoscan and the NVIDIA IGX and NVIDIA Jetson platforms for edge AI and robotics, the Sensor Bridge unlocks new edge-to-cloud applications, enables AI/ML inferencing and facilitates the adoption of AI in medical, industrial, and automotive markets.